Understanding the stability of deep control policies for biped locomotion

被引:3
|
作者
Park, Hwangpil [1 ,3 ]
Yu, Ri [1 ]
Lee, Yoonsang [4 ]
Lee, Kyungho [5 ]
Lee, Jehee [2 ]
机构
[1] Seoul Natl Univ, Seoul, South Korea
[2] Seoul Natl Univ, Dept Comp Sci & Engn, Seoul, South Korea
[3] Samsung Elect, Suwon, South Korea
[4] Hanyang Univ, Comp Sci, Seoul, South Korea
[5] NC Soft, Sungnam, South Korea
来源
VISUAL COMPUTER | 2023年 / 39卷 / 01期
关键词
Biped locomotion; Deep reinforcement learning; Gait analysis; Physically based simulation; Push-recovery stability; RECOVERY;
D O I
10.1007/s00371-021-02342-9
中图分类号
TP31 [计算机软件];
学科分类号
081202 ; 0835 ;
摘要
Achieving stability and robustness is the primary goal of biped locomotion control. Recently, deep reinforcement learning (DRL) has attracted great attention as a general methodology for constructing biped control policies and demonstrated significant improvements over the previous state-of-the-art control methods. Although deep control policies are more advantageous compared with previous controller design approaches, many questions remain: Are deep control policies as robust as human walking? Does simulated walking involve strategies similar to human walking for maintaining balance? Does a particular gait pattern affect human and simulated walking similarly? What do deep policies learn to achieve improved gait stability? The goal of this study is to address these questions by evaluating the push-recovery stability of deep policies compared with those of human subjects and a previous feedback controller. Furthermore, we conducted experiments to evaluate the effectiveness of variants of DRL algorithms.
引用
收藏
页码:473 / 487
页数:15
相关论文
共 50 条
  • [1] Understanding the stability of deep control policies for biped locomotion
    Hwangpil Park
    Ri Yu
    Yoonsang Lee
    Kyungho Lee
    Jehee Lee
    The Visual Computer, 2023, 39 : 473 - 487
  • [2] Push-Recovery Stability of Biped Locomotion
    Lee, Yoonsang
    Lee, Kyungho
    Kwon, Soon-Sun
    Jeong, Jiwon
    O'Sullivan, Carol
    Park, Moon Seok
    Lee, Jehee
    ACM TRANSACTIONS ON GRAPHICS, 2015, 34 (06):
  • [3] Contribution to the integrated control of biped locomotion mechanisms
    Vukobratovic, Miomir K.
    Rodic, Aleksandar D.
    INTERNATIONAL JOURNAL OF HUMANOID ROBOTICS, 2007, 4 (01) : 49 - 95
  • [4] Combined control of CPG and torso attitude control for biped locomotion
    Takahashi, M
    Narukawa, T
    Miyakawa, K
    Yoshida, K
    2005 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS, VOLS 1-4, 2005, : 346 - 351
  • [5] An empirical exploration of a neural oscillator for biped locomotion control
    Endo, G
    Morimoto, J
    Nakanishi, J
    Cheng, G
    2004 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION, VOLS 1- 5, PROCEEDINGS, 2004, : 3036 - 3042
  • [6] Towards the Enhancement of Biped Locomotion and Control Techniques - Walking Pattern Classification
    Yuksel, Basak
    Leblebicioglu, Kemal
    PROCEEDINGS OF THE SIXTEENTH INTERNATIONAL SYMPOSIUM ON ARTIFICIAL LIFE AND ROBOTICS (AROB 16TH '11), 2011, : 918 - 921
  • [7] Biped Locomotion Control through a Biomimetic CPG-based Controller
    Santos, Cristina P.
    Alves, Nuno
    Moreno, Juan C.
    JOURNAL OF INTELLIGENT & ROBOTIC SYSTEMS, 2017, 85 (01) : 47 - 70
  • [8] Biped Locomotion Control through a Biomimetic CPG-based Controller
    Cristina P. Santos
    Nuno Alves
    Juan C. Moreno
    Journal of Intelligent & Robotic Systems, 2017, 85 : 47 - 70
  • [9] AM-RPPO Based Control Method for Biped Adaptive Locomotion
    Ma L.
    Liu C.
    Lin L.
    Xu B.
    Chen Q.
    Jiqiren/Robot, 2019, 41 (06): : 731 - 741
  • [10] Adaptive Asymptotic Stable Biped Locomotion
    Liu, Li Michael
    Liang Wenyuan
    2014 33RD CHINESE CONTROL CONFERENCE (CCC), 2014, : 8375 - 8380